Perinatal Obsessive–Compulsive Disorder: Epidemiology, Phenomenology, Etiology, and Treatment

Abstract

Purpose of Review

We review recent evidence concerning the epidemiology, etiology, and treatment of obsessive–compulsive disorder (OCD) in the perinatal period. We examine studies reporting on rates of both new-onset OCD and exacerbation in both pregnancy and postpartum; explore both biological and psychosocial risk factors for the disorder; and review the latest evidence concerning treatment.

Recent Findings

Evidence is limited in all areas, with rates of both OCD and subthreshold obsessive–compulsive symptoms varying widely across studies. Prevalence is likely higher in the perinatal period than in the general population. Clinical features in the perinatal period are more likely than at other times to concern harm to the child, with contamination and aggressive obsessions and cleaning and checking compulsions especially common. Research into the biological etiology is too limited at this time to be definitive. Both observational and randomized controlled trials support cognitive behavioral therapy with exposure and response prevention (CBT with ERP) as a first-line treatment, with limited evidence also supporting the use of selective serotonin reuptake inhibitors (SSRIs). Treatment considerations in the perinatal period must weigh the risks of treatment vs. the risks of untreated illness.

Summary

Perinatal OCD is common and can be impairing. Clinical features differ somewhat compared to non-perinatal periods. Treatment does not differ from that used in the general population, though evidence pertaining specifically to the perinatal period is sparse.

Introduction

Obsessive–compulsive disorder (OCD) is a common, sometimes debilitating, and under-recognized psychiatric illness, affecting approximately 1–2% of the population (and a larger group with sub-threshold symptoms) [1–3]. Clinical features include unbidden intrusive thoughts or images (obsessions), often accompanied by mental or physical rituals designed to lessen the distress of the obsessions. OCD shares with anxiety disorders a higher prevalence among women compared to men (about 1.6 times) [3, 4], and considerable research points to increased incidence and prevalence at times of reproductive transition [5–7]. The perinatal period is a time of especially increased vulnerability, not only for the disorder, but also for sub-threshold symptoms [8–11]. In one study, fully 65% of new parents reported experiencing intrusive thoughts, though for many these were fleeting and did not affect function [5, 12]. In addition, obsessions are a common feature of other mood and anxiety disorders arising in the perinatal period [13].

An understanding of perinatal OCD is a critical skill for clinicians who treat pregnant and postpartum women. Perinatal OCD has adverse effects for both mother and child. Pregnancies affected by OCD have a higher risk for adverse obstetrical and neonatal outcomes (including gestational hypertension, preeclampsia, poor fetal growth, preterm birth, and elevated inflammatory markers in the neonate) [14–17]. Women with perinatal OCD report poorer quality of life, greater distress, and lower maternal responsiveness [18, 19].

While the disorder resembles OCD at other points in the lifetime, there are some unique clinical features, and evaluation that does not account for these nuances will often miss OCD and attribute symptoms instead to mood or anxiety disorders, with which there are crucial differences in treatment. In addition, the intrusive thoughts of OCD can be mistaken for the delusions of postpartum psychosis, leading either to a failure to recognize the latter (a psychiatric emergency) or, more usually, to a misinterpretation and overreaction to OCD, which can lead to inappropriate actions that may include a higher level of care than is warranted and/or involvement of child services.

This article offers a brief overview of the current understanding of perinatal OCD, focusing on epidemiology and phenomenology, etiology, and treatment.

Epidemiology and Phenomenology

While exact numbers differ across studies, there is considerable evidence that the perinatal period is a time of risk for both new-onset OCD and exacerbation of existing OCD. Rates for new-onset OCD in pregnancy range from 2 to 22%, [2, 11, 20–25], with evidence indicating somewhat higher risk in the postpartum (from 2 to 24%) [8–10, 26]. In addition, studies have shown that 8–70% of women with existing OCD will experience perinatal exacerbations [7, 25, 27–30]. A recent article that carefully characterized women using strict diagnostic criteria found a period prevalence of 7.8% in pregnancy and 16.9% in the postpartum [31].

The clinical features of perinatal OCD have both similarities to and differences from OCD in the general population. As with general OCD, the disorder is characterized by obsessions (intrusive thoughts, ideas, images or impulses), compulsions (repetitive or ritualistic behaviors), or both. To meet criteria for the disorder (as opposed to subthreshold symptoms), the symptoms must affect functioning by being very time-consuming or causing clinically significant distress, and cannot be explained by a different condition [32]. Unique to the perinatal period, obsessions often center on infant harm; aggressive intrusive thoughts of infant harm are especially common in the postpartum [5, 7, 33]. Some literature shows that obsessions that begin in pregnancy are likely to focus on contamination and can be accompanied by cleaning and washing compulsions, while postpartum onset obsessions are more likely to focus on infant harm and be accompanied by checking and avoiding compulsions [33–35].

Thoughts of infant harm in particular can be graphic (including violent or sexual images) and extremely distressing. They cause acute suffering in the mother, and many women are reluctant to disclose such thoughts to a healthcare provider. It is therefore critically important for clinicians to ask specifically and non-judgmentally about intrusive thoughts. It is also of critical importance that clinicians understand how to distinguish between these ego-dystonic intrusive thoughts (which are horrifying to the mother) and the ego-syntonic thoughts of infant harm that can be a feature of postpartum psychosis [36].

Compulsions in the perinatal period are mostly likely to take two forms, in keeping with the most common types of obsessions: contamination-related rituals (such as cleaning and sterilizing baby equipment) and checking or ordering rituals (such as repeated checking on a sleeping infant), with checking the most common perinatal compulsion [33]. If obsessions are extreme, they can also lead to avoidance—for example, the woman who fears dropping her child out the window may avoid all windows, and the woman who has images of her child being stabbed may refuse to use knives or enter the kitchen. Most compulsions and avoidances do not directly harm the child, but in extreme cases, they can— for example, the woman who cleans her child too vigorously or using inappropriate household cleaning products, or the woman who so avoids her child that neglect is the result. Even when symptoms are not extreme, however, they can have deleterious effects on both mother and child by impairing mother–child bonding and attachment [18]. Impairments in attachment can, of course, affect child temperament and fussiness in the short term and can also lead to a lifelong vulnerability to psychiatric disorders.

Perinatal OCD is often comorbid with other psychiatric disorders, and it is important for clinicians to understand how to distinguish a mood or anxiety disorder (possibly with sub-threshold obsessions) from OCD, as treatment differs. The most common mis-diagnoses are generalized anxiety disorder (GAD) and perinatal depression. While both disorders can present with worry and rumination, which can be confused with obsessions, the content of such thoughts in GAD and depression is generally tied to reallife circumstances; in other words, exaggerated levels of distress surrounding things that are actually happening or will happen. Such worries and ruminations generally shift over time. In perinatal OCD, by contrast, intrusive thoughts center on senseless and irrational events and do not shift over time [37].

While it is important to distinguish perinatal OCD from mood and anxiety disorders because of differences in treatment, it is even more important to distinguish it from the psychiatric emergency postpartum psychosis, which can also present with thoughts of infant harm. Women with OCD have preserved insight, are extremely distressed by their thoughts, and will protect or even avoid their infants to prevent harm. Women with postpartum psychosis, by contrast, have delusional beliefs without insight [18, 38, 39]. This crucial distinction means that women with postpartum psychosis are at elevated risk of infant harm, while those with OCD are not [40]. Women with postpartum psychosis nearly always require inpatient hospitalization and alternate care for their infants, and the involvement of social services to ensure the protection of the child is often necessary. Women with OCD, on the other hand, are more appropriately treated in outpatient or partial hospital settings that do not require separation from their babies, and it is rarely necessary to involve social services for the protection of the child (see Table 1).

Table 1.

Thoughts of intrusive harm to the child: distinguishing between obsessions and delusions

Obsessions	Delusions
Intrusive thoughts are unwanted and extremely distressing (ego-dystonic)	Intrusive thoughts may be unbidden but do not cause significant distress (ego-syntonic)
Can be sexual, religious, or violent	Can also be sexual, religious, or violent; content often bizarre or unusual
Patient has no desire to act on thoughts	Patient may want to or feel compelled to act on thoughts
Patient may engage in avoidance or in checking/reassuring compulsions to ease distress	No compulsions
Preserved insight	Poor insight, distortion of reality
Example: Mother has intrusive thoughts about throwing baby out the window. She locks all windows, closes blinds, repeatedly checks locks, and refuses to go to the side of the room where windows are.	Example: Mother believes child has sinned and that God has commanded mother to drop child out the window to punish this sin.

Etiology

The etiology of perinatal OCD at this time is poorly understood, but it is clear that there are both psychosocial and biological contributors. Some studies report that obstetric complications and mode of delivery may affect risk [2, 14, 41–43], but this is not a consistent finding. A personal or family history of mood disorders increases risk [27, 44], as does the presence of dysfunctional obsessive beliefs or an avoidant or obsessive–compulsive personality [5, 23, 45]. Primiparity is a known risk factor and may be due in part to biological influences and in part to the new responsibilities associated with the role transition of new motherhood [45].

Research into the biological etiology of perinatal OCD is very much in its infancy. A large-scale multi-site trial currently underway (NIMH R01MH11824) may, in a few years, increase our understanding of the biology. In the meantime, we must look to the overlap between research on general OCD and that on perinatal mood and anxiety disorders to come up with some plausible hypotheses about the biological origins of perinatal OCD. Genetic contributions have been examined in both general OCD and perinatal depression in the genes for MAO-A, BDNF, and the estrogen and oxytocin receptors, but with no clear consistent findings [46]. Polymorphisms in two genes (5-HTTLPR and COMT) have more consistently been implicated in both general OCD and perinatal depression and so may have more applicability to perinatal OCD [46–48]. Epigenetic biomarkers have been established that predict the exacerbation of OCD symptoms during pregnancy, and this remains a promising future area of research [49, 50].

The immune system is one plausible area of research for the biological etiology of perinatal OCD. There are substantial immune changes in even healthy pregnancy, to prevent the rejection by the mother of the fetus, and immune dysregulation has been implicated in various perinatal morbidities (including mood and anxiety disorders) as well as in general OCD [51–54]. Due to inconsistencies in measurement tools (both of psychiatric diagnoses and of immune markers), however, it is too early to say definitively whether immune dysregulation will prove a strong component.

The same inconclusive conclusion must be applied at this stage to other biological systems. Dysregulation of several neurotransmitter systems, including serotonin, glutamate, and GABA, has also been suggested for both general OCD and perinatal mood and anxiety disorders, making these systems possible targets for future research in perinatal OCD [55–60]. The stress response system may also be involved, with some research showing that women with postpartum OCD have higher levels of basal salivary cortisol than do healthy postpartum controls [61]. Finally, reproductive and other hormones (including estrogen, progesterone, thyroid hormone, and oxytocin) are also ripe areas for future research in perinatal OCD, as all have been examined in perinatal mood and anxiety disorders.

Sleep deprivation may also prove to be a promising future avenue of research. General OCD has been associated with disturbed sleep [62, 63]. All postpartum women experience significant sleep disruption, and those vulnerable to OCD may be more sensitive to such disruption [10, 64–66]. Postpartum obsessions involving fear of harm coming to the baby may make it difficult for women to sleep even when they have the chance, while postpartum compulsions such as checking and cleaning may use up time that could otherwise be spent sleeping [64, 67].

Treatment

As OCD symptoms can be debilitating and can impair the functioning of the entire family, it is vital to recognize and treat them. Unfortunately, there is little research on treatment of OCD specific to the perinatal period, and there are no specific treatment guidelines. There is no reason to expect that perinatal OCD will respond differently to treatment than general OCD, but without more research specific to this population, we cannot be certain.

In the perinatal period, treatment requires a careful consideration of the risks of treatment compared to the risks of no treatment. The risks attaching to psychotropic drugs have been studied extensively at this point, and while there are no randomized controlled trials to guide our treatment choices there is a large and largely reassuring body of observational literature. Studies that use appropriate control groups show few (though not absent) risks associated with the medications most commonly used for OCD. It is imperative for clinicians to view this information in the light of known information about the risks to mother, child, and family of NOT treating the illness. Although there is little research specific to the risks of untreated perinatal OCD, we know that perinatal mood and anxiety disorders, as well as other types of stress and distress in pregnancy, are associated with substantial risk, including relapse of psychiatric illness for the mother; preterm birth; and deficits in cognitive and emotional development of the child [68–72]. It is therefore prudent to use for the perinatal OCD patient a treatment approach that best minimizes risk—by using the lowest effective dose of as few medications as possible. In other words, if it is possible to get the mother well without medication, do that—but if not, treat her and do not be stingy with dosage. Treating her with lower doses than are effective just exposes the baby to both the illness and the medications, while treating to remission creates FEWER exposures for the baby (see Fig. 1).

Fig. 1.

Treatment algorithm for perinatal OCD. From Hutner et al. eds., The APA Textbook of Women’s Reproductive Mental Health (Washington, DC: APA Publishing, 2021)

Nonpharmacological Approaches

Cognitive behavioral therapy (CBT) is the only therapy with evidence to support its effectiveness in OCD [73]. CBT with exposure and response prevention (ERP) has the most evidence [74–76] and should be considered the first-line treatment. While numerous randomized-controlled trials support the use of CBT with ERP in the general population, evidence for the perinatal period specifically is largely limited to observational trials, all of which report improvement in a limited number of patients without a comparator group [77–81]. One recent trial randomized patients with postpartum OCD to intensive CBT vs. treatment as usual and found a significant reduction in OCD symptoms in the intervention group [82]. Treatment guidelines from the general population can be applied in perinatal OCD, with 13–20 weekly sessions followed by a 3–6-month period of booster sessions.

While CBT with ERP is the only evidence-based psychotherapeutic approach, other forms of therapy can be useful for some of the symptoms of perinatal OCD. Family therapy, supportive therapy, motivational interviewing, and dialectical behavioral therapy can be helpful at alleviating distress caused by symptoms, especially for patients who are unable to engage in CBT with ERP due to the severe distress caused by their symptoms [73]. These are, however, stop-gap measures—and the patient who does not benefit from or cannot engage in CBT with ERP will likely require psychopharmacological treatment.

Pharmacological Treatment

The FDA has approved the selective serotonin reuptake inhibitors (SSRIs) fluoxetine, fluvoxamine, paroxetine, and sertraline as well as the tricyclic antidepressant (TCA) clomipramine for OCD, and all are considered first-line agents [32] that are compatible with pregnancy and lactation. Escitalopram and citalopram are also efficacious and can be used despite the lack of FDA indication [83–85]. Monotherapy with one of these agents should be attempted first, and as always in the perinatal period, it is important to maximize one drug before adding a second, to reduce the number of exposures for the fetus or breastfeeding infant. As OCD generally requires higher doses than mood and anxiety disorders—and as pharmacokinetic changes in pregnancy result in plasma concentration drops of 40–50% by the end of pregnancy—doses may need to be much higher than most clinicians are accustomed to in order to be efficacious. Modest evidence also supports second-line approaches, including augmentation with antipsychotics and venlafaxine and mirtazapine monotherapy [86–92].

FDA approval and evidence for the treatments listed above is based on non-perinatal patients, and only open-label trials exist during the perinatal period; efficaciousness in these few trials is largely consistent with that from the general population [44, 79, 93–95]. As with the general population, treatment should be continued for 1–2 years before considering a gradual taper [73]. There is no reason to alter this in the perinatal period, as the relatively benign risks and side effect profiles of the approved drugs make them compatible with pregnancy and breastfeeding.

Other Somatic Interventions

Transcranial magnetic stimulation, deep-brain stimulation and other neurosurgical approaches, and electroconvulsive therapy have all been investigated for general OCD. The significant morbidity associated with these, as well as the small amount of evidence supporting their efficacy (including none in the perinatal period), makes them inadvisable for pregnant and postpartum women in all but the most egregiously treatment-refractory cases.

Conclusions

OCD in the perinatal period is common and often misdiagnosed. Incidence and prevalence rates differ widely across studies, but most literature indicates at the least a modest increase in risk during pregnancy and especially postpartum. While clinical features can resemble those of OCD in the general population, symptoms are often centered around thoughts of infant harm (either direct or indirect), and at their most severe can induce significant distress in mothers and at times lead to avoidance or even neglect of the child. Etiology is poorly understood, though prior history of mood or anxiety disorders and certain personality types enhance risk. There may be both genetic and epigenetic risk factors, and growing but still quite limited evidence points to dysregulation in the stress response system, the immune system, and the neuroendocrine system. Women with OCD rarely act on intrusive thoughts of child harm (unlike those with postpartum psychosis), and most can be appropriately treated on an outpatient basis. CBT with ERP and SSRIs are both first-line treatments, but few data exist on unique features of treatment in the perinatal period. Research into all aspects of this common and debilitating disorder has lagged behind those of other perinatal mood and anxiety disorders.